ARCH GARCH Assignment
ARCH GARCH Assignment
ARCH/GARCH Model
Sruti Suhasaria
212pge020
Introduction:
The autoregressive conditional heteroscedasticity (ARCH) model is a statistical
model for time series data that describes the variance of the current error
term or innovation as a function of the actual sizes of the previous time
periods' error terms.
The generalized autoregressive conditional heteroscedasticity (GARCH) process-
is an econometric term used to describe an approach to estimate volatility in
financial markets.
Data & Methodology:
The data used in this assignment is a time series data of ICICI Prudential Life
Insurance Company Ltd from finance industry which was extracted from
website Yahoo Finance WIPRO where I have used NSE’s real time price and
found the independent variable OPEN.
Result & Interpretation:
The first step is to find which series this data set belongs to through the ADF
test:
ADF Test:
Null Hypothesis: D(OPEN) has a unit root
Exogenous: Constant, Linear Trend
Lag Length: 0 (Automatic - based on SIC, maxlag=15)
t-Statistic Prob.*
Here, we reject the null hypothesis and find that the dataset is an I (1) series.
But, then we cannot model an I (1) series using GARCH model as our dataset
should be stationary.
Hence, an open return must be created using log, and we get the below graph
which clearly indicates it is a stationary series.
2. To model mean & variance, we use the ARCH/GARCH model due to the
presence of an ARCH effect-
From the above table we see that in the Variance equation the ARCH term is
INSIGNIFICANT.
B) Now, we add the GARCH term-
Standard Deviation
Variance
Log(VAR)
E) Now, by trying E-GARCH and removing the GARCH term and by adding an
Asymmetric term in the equation it was seen that the absolute of the residuals
is SIGNIFICANT.
F) Now, by coming back to the GARCH model and adding abs(resid) to the
variance regressors and keeping only ARCH value
Here we see that since all the values are above zero and positive, we can use
the GARCH model for this time series dataset.